Technical Report HCSU-019

EFFICACY OF A COMMERCIAL FOR PROTECTING HAWAI`I `AMAKIHI FROM FIELD ISOLATES OF

Carter T. Atkinson1, Kimberly C. Wiegand2, Dennis Triglia3, and Susan I. Jarvi2

1 U.S. Geological Survey, Pacific Island Ecosystems Research Center, Kilauea Field Station, P.O. Box 44, Hawaii National Park, HI 96718 2Department of Pharmaceutical Sciences, College of Pharmacy, University of Hawai`i, Hilo, 200 W. Kawili Street, Hilo, HI 96720 3 Hawai‘i Cooperative Studies Unit, University of Hawai‘i at Hilo, Pacific Aquaculture and Coastal Resources Center, P.O. Box 44, Hawai‘i National Park, HI 96718

Hawai‘i Cooperative Studies Unit University of Hawai‘i at Hilo Pacific Aquaculture and Coastal Resources Center (PACRC) 200 W. Kawili St. Hilo, HI 96720 (808) 933-0706

The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the opinions or policies of the U.S. Government. Mention of trade names or commercial products does not constitute their endorsement by the U.S. Government.

Technical Report HCSU-019

EFFICACY OF A COMMERCIAL CANARYPOX VACCINE FOR PROTECTING HAWAI`I `AMAKIHI FROM FIELD ISOLATES OF AVIPOXVIRUS

Carter T. Atkinson1, Kimberly C. Wiegand2, Dennis Triglia3, and Susan I. Jarvi2

1U.S. Geological Survey, Pacific Island Ecosystems Research Center, Kilauea Field Station, P.O. Box 44, Hawai`i National Park, HI 96718, USA 2Department of Pharmaceutical Sciences, College of Pharmacy, University of Hawai`i at Hilo, 200 W. Kawili Street, Hilo, HI 96720, USA 3Hawai`i Cooperative Studies Unit, University of Hawai`i at Hilo, Pacific Aquaculture and Coastal Resources Center, P.O. Box 52, Hawai`i National Park, HI 96718, USA

KEY WORDS

Avipoxvirus, avian pox, vaccine, Biomune Poximmune C®, Hawaiian forest birds, honeycreeper, Hawai`i `Amakihi, Hemignathus virens

CITATION Atkinson, C.T., K.C. Wiegand, D. Triglia, and S.I. Jarvi. 2010. Efficacy of a commercial canarypox vaccine for protecting Hawai`i `Amakihi from field isolates of Avipoxvirus. Hawai`i Cooperative Studies Unit Technical Report HCSU-01. University of Hawai`i at Hilo. 47 pp., incl. 23 figures. Hawaii Cooperative Studies Unit University of Hawai`i at Hilo Pacific Aquaculture and Coastal Resources Center (PACRC) 200 W. Kawili St. Hilo, HI 96720 (808)933-0706

September 2010

i

This product was prepared under Cooperative Agreement CA03WRAG0036 for the Pacific Island Ecosystems Research Center of the U.S. Geological Survey

ii

Table of Contents

LIST OF FIGURES ...... iii

ABSTRACT ...... 1

INTRODUCTION...... 2

METHODS ...... 3

Relationship of Vaccine to Wild Pox Isolates from Hawaiian Birds: ...... 3 Vaccination and Challenge with Live : ...... 4 Biosecurity: ...... 6 Statistical Analysis: ...... 7

RESULTS ...... 7

Analysis and Classification of Vaccine Virus: ...... 7 Vaccination:...... 7 Challenge with Pox Variant 2 ...... 9 Challenge with Pox Variant 1 ...... 10 Challenge with Fowlpox ...... 12

DISCUSSION ...... 12

ACKNOWLEDGMENTS ...... 16

LITERATURE CITED ...... 17

ii

LIST OF FIGURES Figure 1. A neighbor-joining tree (Kimura 2-parameter corrected distances) showing relationships among pox virus sequences amplified from native and non-native birds and ® Poximmune C (Vaccine C06)...... 19

Figure 2. Normal vaccine “take” in Hawai`i `Amakihi 321 that resolved by Day 34 PI. .20

Figure 3. Duration of vaccine lesions in 15 Hawai`i `Amakihi that were vaccinated in the ® wing web with Poximmune C ...... 21

Figure 4. Necrotic lesion that developed on Hawai`i `Amakihi 304 after vaccination in ® the wing web with Poximmune C ...... 22

Figure 5. Necrotic wing lesion (arrow) that developed on Hawai`i `Amakihi 305 after ® vaccination with Poximmune C ...... 23

Figure 6. Development of proliferative lesions on Hawai`i `Amakihi 302 after ® vaccination in the wing web with Poximmune C ...... 24

Figure 7. Development of proliferative, necrotic lesions (arrows) on Hawai`i `Amakihi ® 315 after vaccination with Poximmune C ...... 25

Figure 8. Cumulative food consumption (top) and weight change (bottom) for Hawai`i ® `Amakihi vaccinated with Poximmune C ...... 26

Figure 9. Development of proliferative, necrotic lesions on Hawai`i `Amakihi 321 after challenge with Pox Variant 2...... 27

Figure 10. Kaplan Meier Survival Plot of vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 2...... 28

Figure 11. Cumulative food consumption (top) and weight change (bottom) among vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 2...... 29

Figure 12. Hematoxylin and eosin-stained section of a wing lesion from Hawai`i `Amakihi 309 after vaccination and challenge with Pox Variant 2...... 30

Figure 13. Wrinkled, edematous abdominal skin (upper inset) and hematoxylin and eosin-stained section of abdominal skin from Hawai`i `Amakihi 307 after challenge with Pox Variant 2...... 31

Figure 14. Hematoxylin and eosin stained section of the subdermal region of Hawai`i `Amakihi 331...... 32

iii

Figure 15. Development of small lesion in Hawai`i `Amakihi 328 after vaccination with ® Poximmune C and challenge with Pox Variant 1 ...... 33

Figure 16. Development of a moderate lesion in Hawai`i `Amakihi 312 after vaccination ® with Poximmune C and challenge with Pox Variant 1...... 34

Figure 17. Unvaccinated control Hawai`i `Amakihi 323 inoculated with Pox Variant 1. 35

Figure 18. Unvaccinated control Hawai`i `Amakihi 78481 inoculated with Pox Variant 1...... 36

Figure 19. Kaplan Meier Survival Plot of vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 1...... 37

Figure 20. Cumulative food consumption and weight change for vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 1 ...... 38

Figure 21. Small lesion (arrow) in the wing web of Hawai`i `Amakihi 302 after challenge with Fowlpox virus...... 39

Figure 22. Weight change and cumulative food consumption for vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 1...... 40

Figure 23. Kaplan Meier Survival Plot of vaccinated and unvaccinated Hawai`i `Amakihi challenged with Fowlpox...... 41

iv

ABSTRACT

At least three variants of avian pox virus are present in Hawai’i - Fowlpox from domestic poultry and a group of genetically distinct that cluster within two clades (Pox Variant 1 and Pox Variant 2) that are most similar to Canarypox based on DNA sequence of the virus 4b core protein gene. We tested whether Hawai’i ‘Amakihi can be protected from wild virus isolates with an attenuated live Canarypox vaccine that is closely related to isolates that cluster within clade 1 (Pox Variant 1) based on sequence of the attenuated Canarypox virus 4b core protein. Thirty-one (31) Hawai`i ‘Amakihi (Hemignathus virens) with no prior physical evidence of pox infection were collected on Mauna Kea from xeric, high elevation habitats with low pox prevalence and randomly divided into two groups. One group of 16 was vaccinated with Poximmune C® while the other group received a sham vaccination with virus diluent. Four of 15 (27%) vaccinated birds developed potentially life-threatening disseminated lesions or lesions of unusually long duration, while one bird never developed a vaccine-associated lesion or “take”. After vaccine-associated lesions healed, vaccinated birds were randomly divided into three groups of five and challenged with either a wild isolate of Fowlpox, a Hawai`i `Amakihi isolate of a Canarypox-like virus from clade 1 (Pox Variant 1) or a Hawai`i `Amakihi isolate of a Canarypox-like virus from clade 2 (Pox Variant 2). Similarly, three random groups of five unvaccinated ‘Amakihi were challenged with the same virus isolates. Vaccinated and unvaccinated ‘Amakihi challenged with Fowlpox had transient infections with no clinical signs of infection. Mortality in vaccinated ‘Amakihi that were challenged with Pox Variant 1 and Pox Variant 2 ranged from 0% (0/5) for Pox Variant 1 to 60% (3/5) for Pox Variant 2. Mortality in unvaccinated ‘Amakihi ranged from 40% (2/5) for Pox Variant 1 to 100% (5/5) for Pox Variant 2. While the vaccine provided some protection against Pox Variant 1, serious side effects and low efficacy against Pox Variant 2 make it risky to use in captive or wild honeycreepers.

1

INTRODUCTION

Introduced -borne avian pox and malaria continue to have serious effects on the long term recovery of Hawai`i’s endemic forest birds. Both diseases negatively impact wild populations (Atkinson et al. 1995, Vanderwerf 2001, Atkinson and LaPointe 2009) and pose a serious threat to long term recovery of threatened and endangered honeycreepers (U. S. Fish and Wildlife Service 2006). Few practical methods exist for control of pox or malaria in wild populations and this problem has made translocation or introduction of captive reared birds into habitats that would otherwise be suitable for recovery very risky. A vaccine that provides protection to one or both of these diseases may help released birds survive longer and increase the odds that they can establish breeding populations. Infections with Avipoxvirus can occur in three forms – cutaneous, wartlike nodules on unfeathered skin around the eyes, beak and feet, a diphtheritic form of infection on mucous membranes of the mouth and upper respiratory tract, and a rare systemic form that may occur throughout tissues of the infected host (van Riper and Forrester 2007). Depending upon the placement and number of lesions, infected birds may encounter difficulty seeing, feeding, breathing or perching. Pox virus infections in animals are often immunosuppressive (Smith and Kotwal 2002) and co-infections of avian pox and malaria act synergistically in Hawaiian honeycreepers under experimental conditions to increase malarial parasitemia and mortality (C. Atkinson, unpublished data). We have propagated over 20 viral isolates in tissue culture that were obtained from native and non-native forest birds on Oahu, Maui, Molokai and Hawaii Islands. There are at least three different strains or variants of the virus among the samples we have analyzed to date based on genetic analysis of both a 538 bp and a shorter 116 bp fragment of the conserved 4b core protein gene of the virus (Jarvi et al. 2008). One type, isolated from naturally infected domestic chickens in Volcano on Hawaii Island, is identical to Fowlpox isolates from other parts of the world. The two other variants differ significantly from the Fowlpox isolates and appear to be specific to . One of the two variants clusters closely with Canarypox isolates in phylogenetic analyses while the other appears unique to Hawai`i (Jarvi et al. 2008).

2

We tested the efficacy of a commercially available, live-attenuated Canarypox vaccine (Biomune Poximmune C®) against challenge with isolates of three variants of virus we have detected in forest bird populations, particularly the non-Fowlpox-like types that we have only isolated from hosts. Our goal was to evaluate a commercial vaccine to see if it can be used as a tool for protecting threatened and endangered captive- reared or wild honeycreepers during release or translocation or for protecting critically endangered wild populations that face significant threats from disease transmission.

METHODS Relationship of Vaccine to Wild Pox Isolates from Hawaiian Birds: Biomune Poximmune C® (Ceva Biomune, Lenexa, KS) is a live attenuated Canarypox vaccine that is approved for use in canaries and other small passerine birds. We initially established the relationship of the attenuated vaccine virus to isolates of the three variants of poxvirus that have been isolated from native and non-native forest birds in Hawaii. DNA was extracted from lyophilized vaccine using the DNeasy® Tissue Kit (QIAGEN, Valencia, CA) following manufacturer’s protocols. A 116 bp portion of the highly conserved virus 4b core protein was amplified by PCR using primers PV4B.P4 (5’- CACATGTTAAGGGGTCTCTATC-3’) and PV4B.P5 (5’- TGTAGTATCAATAAGCGCTTGGT-3’). Three microliters (μL) of extracted DNA template was used in 100 μL PCR reactions containing 1X reaction buffer (Roche Diagnostics, Penzber, DE), 0.8 mM of total dNTP, 1.25 units of Taq polymerase (Roche

Diagnostics), 1 mM MgCl2 and 0.8 μM of primers PV4B.P4 and PV4B.P5. Samples were subjected to an initial denaturation step of 3 min at 94oC followed by 40 cycles of denaturation for 30 sec at 94oC, annealing for 1 min at 53oC, and extension for 1 min at 72oC in a MJ Research PTC-100 thermocycler with a heated lid (MJ Research, Ramsey, MN). Products from the PCR were isolated by gel extraction from a 1.5% agarose gel. Because of the proofreading capabilities of Roche Taq, addition of an A overhang was necessary before cloning. Briefly, 8.6 μL of PCR product was used in 50 μL reactions

containing 1X PCR buffer, 2.5 mM MgCl2, 5 mM dATP, 1 unit Taq (Promega, Madison, WI) and incubated at 72C for 15 min. PCR products were then directly cloned using the TOPO®-TA Cloning® Kit

3

(Invitrogen Corp., Carlsbad, CA) following manufacturer’s protocols. Plasmids were isolated using the QIAprep® Spin Miniprep Kit Protocol (QIAGEN) following manufacturer’s protocol. Plasmids were screened by EcoRI digest to determine if an insert of the correct size was present. Four clones were sequenced in both directions on an ABI sequencer (UH, Manoa). All sequences were proofed and analysed using Sequencher® (GeneCodes Corp, Ann Arbor, MI). We assessed the relationship among pox vaccine sequences and sequences reported in Jarvi et al. (2008) by constructing a neighbor-joining tree from Kimura 2- Parameter corrected distance matrices using MEGA 2.0 (Kumar et al. 2000). Vaccination and Challenge with Live Virus: Thirtyone Hawai`i `Amakihi (Hemignathus virens) of mixed age and sex were captured in xeric, high elevation habitat at Pu`u La`au in Mauna Kea Forest Reserve. This area has an extremely low prevalence of pox infection in resident Hawai`i `Amakihi and prior surveys of over 1000 resident birds reported a prevalence of only 0.4% (5/1250) (van Riper 1991). In addition, birds were examined carefully for active lesions and presence of missing digits which might indicate prior infection with the virus. Hawai`i `Amakihi with no current or prior physical evidence of pox infection were transported to the USGS Aviary in Hawai`i Volcanoes National Park, and acclimated to captivity in a free flight cage. Birds were fed a diet of artificial nectar (Nekton® Nectar Plus), scrambled eggs, and mixed fruits. Eight weeks after capture, 16 Hawai`i `Amakihi were randomly selected, moved to individual cages in a separate aviary, and vaccinated in the wing web following manufacturer’s instructions. The lyophilized vaccine was reconstituted immediately before use with addition of sterile distilled water and administered with a beveled 16 gauge vaccinator needle, supplied with the vaccine, calibrated to carry approximately 10 μl of solution. The remaining 15 Hawai`i `Amakihi were also moved to individual cages in an adjoining room with an independent ventilation system and sham-vaccinated in the wing web with a beveled vaccinator needle dipped in sterile water. Vaccinated and control birds were weighed twice a week and examined carefully for evidence of a slight swelling or vaccine “take” at the site of inoculation. Necton consumption was measured daily for each bird. Birds were also bled once a week by jugular venipuncture to collect 100 μl of heparized whole blood. Blood

4

samples were centrifuged and packed cells and plasma were separated for later diagnostic analysis. If vaccinated Hawai`i `Amakihi did not develop evidence of a vaccine “take” at the site of inoculation within 8 weeks, they were revaccinated in the wing web with freshly reconstituted vaccine and monitored twice a week for evidence of a “take” at the inoculation site for an additional 8 weeks. Birds that did not develop visible “takes” at the inoculation site after a second round of vaccination were removed from the experiment. Pox lesions from two wild Hawai`i `Amakihi that fell within either the Pox Variant 1 clade (HAAM PV013, Ainahou Ranch) or the Pox Variant 2 clade (HAAM PV020, Ohia Estates) and a domestic chicken from Volcano Village, Hawaii (CHCK784 PV001) were ground using a Wheaton Potter-Elvehjem tissue grinder in 4.5 ml of Hank’s Balanced Salt Solution (HBSS; GIBCO #24020-117) containing 5% (w/v) glycerol (Sigma-Aldrich® Corp, St. Louis, MO), 300 U/L penicillin G sodium, 300 ug/ml streptomycin sulfate and 750 ug/ml amphotericin B(as Fungizone, GIBCO®). The amount of tissue used for initial isolation depended on lesion size and ranged from 5 – 74 mg for the three isolates we used. After grinding, the homogenate was transferred to a 15 mL polypropylene conical centrifuge tube (Corning Inc., Corning, NY) and centrifuged for 30 minutes at 800 x g in a Marathon 6K centrifuge (Thermo Fisher Scientific Inc., Waltham, MA) at room temperature. Five (5) mL of the supplemented HBSS was also centrifuged for subsequent use as a “sham” infection inoculum. After the centrifugation, the supernatant fluids (pox and sham) were carefully transferred to new sterile centrifuge tubes and refrigerated at 2 - 8oC for two hours prior to inoculation of monolayer cultures of Muscovy Duck embryonic fibroblasts (MSDEF). Two and one-half milliliters (2.5 mL) of the clarified pox inocula and 2.5 mL of the HBSS “sham” inoculum were each introduced into a seven-day old confluent T-75 flask of MSDEF containing 22.5 mL of Medium 199 (1X) (GIBCO® #11150-059) with Earle’s salts and 2.2 g/l sodium bicarbonate supplemented with 10% Fetal Bovine Serum (FBS) (GIBCO® #16000-044). o Flasks were returned to a 37 C/7.5% CO2 incubator. The medium was changed three and five days post inoculation. Since our pox culture system using monolayers of MSDEF did not produce typical “pocks” after inoculation of virus that could be used to quantify viral dosages, we qualitatively assessed virus concentration by scoring severity of cytopathology within monolayers of virus-inoculated and sham-inoculated MSDEF using

5

an inverted . Monolayers were harvested at either four days post inoculation (HAAM PV020, Ohia Estates), six days post inoculation (CHCK784 PV001) or seven days post inoculation (HAAM PV013, Ainahou Ranch) when monolayers were beginning to peel off the flask surface and cytopathic effects were scored as 2.5 or higher on a scale of 0-5. Cytopathic effects were not evident in monolayers inoculated with sham supernatant. For harvest, infected monolayers were incubated for 30 minutes at 37oC/ 7.5%

CO2 with HBSS without FBS. The media was then removed and 1.5 mL of sterile filtered Phosphate Buffered Saline (PBS), pH 7.4 was added per flask and frozen. Flasks were subsequently thawed and lysates were collected and centrifuged as described earlier to remove the majority of the cellular debris. The supernatant containing live virus was removed and stored at -70o C prior to use. After vaccination “takes” on wing webs of vaccinated Hawai`i `Amakihi had completely healed, five vaccinated and five sham-vaccinated control birds were randomly selected and moved into a third aviary with an independent ventilation system. The suspension containing live Pox Variant 2 virus was thawed and all 10 birds were challenged sequentially by dipping a beveled needled once into the supernatant solution and piercing the skin of the wing web. Challenged birds were monitored and bled as described previously. Lesion development was monitored twice a week by photography. Birds that died were necropsied and representative pieces of skin, lesion, and all major organs were fixed in 10% buffered formalin, embedded in paraffin, sectioned and stained with hematoxylin and eosin for histopathology. At 16 weeks after initial vaccination or eight weeks after revaccination, the remaining 9 vaccinated and 10 sham-vaccinated Hawai`i `Amakihi were randomly divided into two groups of five and four vaccinated birds and two groups of five sham- vaccinated birds, moved into separate aviaries with independent ventilation systems and challenged in the wing web with a vaccinator needle as described earlier with a suspension of Pox Variant 1 virus (one group of 10) and Fowlpox virus (one group of nine). Birds were monitored and necropsied as described earlier. Biosecurity: We took a range of precautions to prevent mixing of pox virus strains during the

6

experiment. Hawai`i `Amakihi infected with different pox virus isolates were housed in separate rooms with independent ventilation systems to avoid potential cross contamination from aerosols. All caging was washed with hot soapy water, sprayed with a 10% solution of chlorine bleach, and rinsed with water prior to and after use. All food dishes and Nekton® tubes were washed every day with hot soapy water in a commercial dishwasher. Each room also had a separate set of brushes, buckets, and hoses to avoid cross contamination. Personnel wore rubber boots, disposable Tyvek® pants, and a single use laboratory coat, used a footbath containing 10% bleach when entering and leaving individual rooms, and thoroughly washed their hands with soap before entering a new room. Pants were discarded and laboratory coats were washed with bleach in a commercial washer/dryer after use in each room. During vaccination, we also cared for and measured control birds first every day to reduce the chance that virus might move from contaminated to uncontaminated areas. Statistical Analysis: Weight and food consumption data were analyzed with a repeated measures ANOVA using the statistical program Systat®, Version 11 (Systat Corp., Chicago, IL). Survival analyses for birds infected with Variant 1 and Variant 2 were done using a Kaplan Meier Survivorship Analysis with Systat®. Statistical tests were considered significant when P < 0.05.

RESULTS Analysis and Classification of Vaccine Virus: Three of the four partial 4b core protein gene clones produced unambiguous sequences that were 100% identical to each other. One representative sequence was used in a neighbor-joining tree with other previously published sequences to establish relationships (Figure 1). The attenuated Canarypox vaccine virus appears most similar to the ATCC Canarypox sequence as well as field isolates of Avipoxvirus Variant 1 (Figure 1). Vaccination: Approximately 7 days after vaccination with Poximmune C®, Hawai`i `Amakihi developed small, 2-3 mm diameter areas of discoloration or swelling at the inoculation

7

site in the wing web (Figure 2). These swellings or vaccine “takes” increased in diameter over the next two weeks for 6 birds (306, 316, 317, 318, 321, 78404) and then healed within 3-6 weeks after inoculation, leaving no scar (Figure 2). By contrast, four Hawai`i `Amakihi (302, 304, 305, 315) developed, extensive, necrotic wing lesions. Wing lesions on three of these birds (302, 304, 305) took 7-8 weeks to heal (Figure 3). One bird (304) developed a large scar in the wing membrane during the healing process (Figure 4), while a second (305) developed a large necrotic lesion that eventually healed completely without scarring (Figure 5). Two birds (302, 315) developed proliferative lesions around the beak (302, 315), eyelids (315), and also on the legs and feet (315) (Figures 6, 7). Lesions on Hawai`i `Amakihi 315 were particularly bad, leading to loss of function in the wing by six weeks after vaccination, loss of function in the right leg by 11 weeks after vaccination, and complete loss of the right leg by 14 weeks after vaccination (Figure 7). This bird subsequently died approximately 17 weeks after vaccination with persistent active lesions on the beak and remaining leg. Six Hawai`i `Amakihi (308, 312, 313, 324, 327, 328) did not develop vaccine “takes” and were subsequently revaccinated eight weeks after the initial vaccination. Five of these six birds developed “takes” that healed within 4-7 weeks after vaccination. One individual, Hawai`i `Amakihi 313, did not develop a swelling that would indicate successful vaccination, and was dropped from the experiment. Among the 15 Hawai`i `Amakihi that developed clear vaccine takes, 11 developed wing lesions that resolved without evidence of scarring within 7 weeks after vaccination. The remaining 4 birds, however, developed potentially life threatening lesions that exhibited extensive , caused loss of function in a wing or leg, or spread beyond the inoculation site to other parts of the body. None of the control birds developed lesions following a sham-vaccination with distilled water. Among all three groups, cumulative food consumption among birds with large vaccine-associated lesions was slightly lower than cumulative food consumption in control birds and birds with small vaccine takes (Figure 8), but differences among subjects were not significantly different (P= 0.211, F = 1.673, df = 2). There were, however, significant within-subject Time (P<0.0001, F = 4565.685, df = 69) and Treatment*Time interactions (P=0.004, F=1.367, df=138), suggesting that treatment had

8

an effect on individual food consumption over time. Among the three experimental groups, birds with large vaccine-associated lesions lost more weight over time than control birds or birds with small vaccine takes (Figure 8), but differences among subjects were not significantly different (P=0.115, F=2.395, df=2). There were, however, significant within-subject Time (P<0.0001, F = 1083.992, df = 18) and Treatment*Time interactions (P=0.007, F=1.735, df=36), suggesting that treatment had an effect on individual weight change over time. Challenge with Pox Variant 2 After challenge with Pox Variant 2, five Hawai`i `Amakihi that had been vaccinated once with Poximmune C® and five unvaccinated control Hawai`i `Amakihi developed wing lesions as early as nine days after infection. Lesions were first evident as reddish swellings at the inoculation site (Figure 9). Lesions on three of five vaccinated birds (306, 318, 321) continued to grow, eventually leading to an extensive thickening and of the wing web that spread throughout the dermis of inoculated birds (Figure 9). Edematous swelling and wrinkling of the dermis spread, eventually involving skin on the abdomen, thighs, head and neck, and all three of these birds died between 40- 50 days post-challenge. One bird (321) developed proliferative lesions that spread to the foot and beak, eventually leading to loss of circulation in the right leg and loss of the upper beak (Figure 9). The two remaining vaccinated birds (316, 317) developed small swellings at the inoculation site by day 9 after challenge that resolved completely by day 21 after challenge. All five unvaccinated control Hawai`i `Amakihi (307, 309, 319, 322, 331) died between 24 – 30 days after challenge. Lesions were similar to those from the three vaccinated birds, beginning as reddish swellings at the inoculation site by day nine that rapidly led to extensive thickening and inflammation of the wing web and a generalized edematous swelling and wrinkling of the that spread over the entire body before death at three to four weeks after inoculation. While significantly fewer vaccinated birds died after challenge with Pox Variant 2 virus (P = 0.002, X2 = 9.651, df = 1), mortality was still 60% (3/5) (Figure 10). In spite of the high mortality among both vaccinated and unvaccinated Hawai`i `Amakihi, challenge with Pox Variant 2 had no significant effects on cumulative food

9

consumption (P = 0.950, F = 0.052, df = 2) or weight change (P = 0.958, F = 0.043, df = 2) during the first 23 days of the study when all birds could be included in the repeated measures analysis (Figure 11). Within-subject effects for cumulative food consumption were significant for Time (P < 0.0001, F = 3892.722, df = 33), indicating that daily food consumption was consistent throughout the study, however, Treatment*Time interactions were not significant (P =1.000, F = 0.476, df = 66), indicating that day to day variability in cumulative food consumption was not affected by treatment status. There were significant Time (P<0.0001, F=1385.366, df =7) and Treatment*Time (P<0.0001, F=3.121, df =14) interactions for weight change, indicating that there were significant day to day variations in weight change within subjects that were affected by treatment. Most interesting was the increase in weight among unvaccinated Hawai`i `Amakihi between Days 20-30 PI, immediately prior to death (Figure 11). By the end of the experiment at Day 58, two vaccinated birds that were still alive had lower totals for cumulative food consumption and slightly higher weights relative to pre-infection values than uninfected control birds (Figure 11), but these differences were not statistically significant (weight change, P=0.279, T = -1.144, df = 10; cumulative food consumption, P= 0.135, T = 1.625, df = 10). Gross and microscopic lesions were similar among vaccinated and unvaccinated birds that died after challenge with Pox Variant 2. Wing lesions and proliferative lesions that spread to other parts of the body had microscopic lesions typical of pox infections, including proliferation and hypertrophy of epithelial cells, particularly cells of the stratum germinativum (Figure 12). Viral inclusion or Bollinger Bodies were common in these cells and the pink-staining inclusions were either amorphous or cuboidal in shape (Figures 12, 13). Necrotic areas, extensive inflammatory infiltrates, and secondary bacterial and fungal infections were prominent in the dermis and subdermis (Figures 12, 13). Epithelial hyperplasia and intracytoplasmaic Bollinger Bodies were evident in both the stratum germinativum and subdermis of edematous and wrinkled skin from the abdomen of vaccinated and unvaccinated birds (Figure 13). Both extracellular viral and mononuclear cells containing pink-staining inclusion bodies were evident in the subdermis of some birds (Figure 14). Challenge with Pox Variant 1

10

Five revaccinated Hawai`i `Amakihi and five unvaccinated Hawai`i `Amakihi were challenged in the wing web with Pox Variant 1. Four of five revaccinated birds (308, 324, 327, 328), i.e. birds that were revaccinated when they did not develop “takes”, developed small lesions at the inoculation site that quickly resolved and disappeared as early as 12 days after challenge (Figure 15). The fifth revaccinated bird (312) had a wing lesion of moderate size after challenge that did not completely heal until 56 days after exposure (Figure 16). The lesion had no effect on wing function and the bird was otherwise healthy. All five unvaccinated Hawai`i `Amakihi (303, 314, 323, 326, 78481) developed necrotic wing lesions after challenge with Pox Variant 1 that were much larger and more debilitating than lesions from the five vaccinated birds (Figure 17). Hawai`i `Amakihi 78481 developed an extensive thickening and inflammation of the wing web that spread throughout the dermis of the body, leading a generalized edematous swelling and wrinkling of the skin (Figure 18), but this was the only bird among the unvaccinated Hawai`i `Amakihi challenged with Pox Variant 1 to exhibit this type of proliferative lesion. Lesions in 3 other unvaccinated birds spread to the beak, feet, or both, where they formed more typical tumor-like swellings. Two of the unvaccinated birds (323, 78481) ultimately died (Figures 17, 18), but differences in survivorship between vaccinated and unvaccinated groups were not statistically significant (P = 0.134, X2 = 2.242, df = 1) (Figure 19). Gross and microscopic lesions, particularly the edematous, dermal lesions of Hawai`i `Amakihi 78481, were similar to those in Hawai`i `Amakihi that were challenged with Pox Variant 2 (Figures 12-14). Challenge with Pox Variant 1 had no significant effects on food consumption (P = 0.881, F = 0.024, df = 1) or weight (P = 0.555, F = 0.380, df = 1) among vaccinated or unvaccinated Hawai`i `Amakihi (Figure 20). Within-subject effects for cumulative food consumption were significant for Time (P < 0.0001, F = 2968.578, df = 47), indicating that daily food consumption was consistent throughout the study. However, Treatment*Time interactions were not significant (P =0.997, F = 0.511, df = 47), indicating that day to day variability in cumulative food consumption was not affected by treatment status. Similarly, within subject effects for weight change were significant for Time (P<0.0001, F = 687.821, df = 9), indicating that within-subject weight change

11

varied over time, but Treatment*Time interactions were not significant (P=0.877, F = 0.490, df = 9), indicating that day to day variability in weight change was not affected by treatment status. Challenge with Fowlpox Four Hawai`i `Amakihi that were vaccinated once with Poximmune C® (78404, 302, 304, 305) were moved into individual cages in a separate room with independent ventilation with five unvaccinated control Hawai`i `Amakihi and challenged with Fowlpox virus as described earlier. Both vaccinated and unvaccinated control birds developed small lesions at the inoculation site within 5 days after challenge that completely healed by 12 days after challenge (Figure 21). There were no differences between the two groups in lesion size or duration. Challenge with Fowlpox had no significant effects on food consumption (P = 0.889, F = 0.021, df = 1) or weight change (P = 0.165, F = 2.401, df = 1) among vaccinated or unvaccinated Hawai`i `Amakihi (Figure 22). Within-subject effects for cumulative food consumption were significant for Time (P < 0.0001, F = 2880.587, df = 79), indicating that daily food consumption was consistent throughout the study. However, Treatment*Time interactions were not significant (P =1.000, F = 0.129, df = 79), indicating that day to day variability in cumulative food consumption was not affected by treatment status. Similarly, within subject effects for weight change were significant for Time (P<0.0001, F = 258.435, df = 18), indicating that within-subject weight change varied over time, but Treatment*Time interactions were not significant (P=0.983, F = 0.414, df = 18), indicating that day to day variability in weight change was not affected by treatment status. One bird (78404) developed neurological signs, including torticollis and loss of balance on Day 69 PI and was found dead the following day. No gross lesions were evident at necropsy and cause of death was unknown and presumably not related to the experimental infection with pox virus. There were no significant differences in survivorship between vaccinated and unvaccinated birds (P = 0.264, X2 = 1.250, df = 1) (Figure 23). DISCUSSION While the impacts of introduced Avipoxvirus on native Hawaiian forest birds have

12

been widely documented (Warner 1968, VanderWerf 2001, van Riper et al. 2002), relatively little information has been available about the identity and diversity of virus strains in Hawaii or whether commercial pox are sufficiently immunogenic to be protective (Tripathy et al. 2000 and Kim and Tripathy 2006a) provided the first evidence that pox virus isolates from Hawaiian birds differ from Fowlpox, are non-pathogenic in domestic chickens, and are closely related to Canarypox based on sequence of the virion assembly protein gene. A more extensive survey of pox virus isolates from native and non-native birds, seabirds, and domestic poultry from Hawaii based on sequence of the virus 4b core protein was recently completed by Jarvi et al. (2008). Two clades (Variant 1 and Variant 2) from passerines and a third clade that formed a distinct basal cluster containing known Fowlpox isolates were identified. The American Type Culture Collection strain of Canarypox was most similar to viral isolates in clade 1 (Pox Variant 1). Sequence analysis of a 116 bp region from the vaccine strain we used (Poximmune C®) also fell within this clade, suggesting that the attenuated vaccine might be particularly protective against challenge with wild viruses that fell within clade 1. Based on the close relationship between viral isolates that fell within clade 1 and Canarypox (Kim and Tripathy. 2006a, Jarvi et al. 2008), we designed a series of experiments to test both the safety and efficacy of Biomune Poximmune C® against three known strains of Avipoxvirus that have been isolated from native and non-native forest birds and domestic poultry from Hawaii. This experiment also provided the opportunity to obtain information about relative susceptibility of a native honeycreeper to Fowlpox and to obtain additional experimental data about relative pathogenicity of Pox Variants 1 and 2 (Jarvi et al. 2008). For a vaccine to be useful for immunizing wild or captive birds, it should be effective after a single dose against known viral variants that birds may encounter in the field, produce long-lasting immunity, and have minimal side effects. During the first phase of the study, we divided captive Hawai`i `Amakihi into two random groups and vaccinated half with Poximmune C® according to the manufacturer’s instructions and half with distilled water to access success of a single dose. Only 10 of 16 birds (63%) developed “takes” after initial vaccination with Poximmune C®, indicating that they had been successfully inoculated with the attenutated virus. The remaining six birds were

13

revaccinated after eight weeks with a freshly reconstituted vaccine and only five of the six (83%) developed “takes” after a second inoculation with the vaccine. Of significant concern was the development of potentially life threatening, proliferative lesions on four of the 15 birds (27%). These vaccine-related lesions caused significant loss of function in the wing that was inoculated and in the most extreme case (Hawai`i `Amakihi 315), proliferative lesions led to loss of the entire right leg. We also observed some effects on cumulative food consumption and weight change among these four birds (Figure 8). Both the low vaccination success rate (63%) and the development of life-threatening lesions among a significant number of vaccinated birds precludes use of this vaccine in captive honeycreepers or threatened or endangered wild birds. Among Hawai`i `Amakihi that recovered from both normal vaccine “takes” and more extensive proliferative lesions, Poximmune C® provided marginal protection to birds that were challenged with Pox Variant 2, increasing survival times and decreasing mortality among vaccinated birds (Figure 10), but mortality was still unacceptably high at 60% (3/5). The vaccine provided the most protection against challenge with Pox Variant 1, as might be expected given the closer relationship of the attenuated vaccine with wild isolates in this clade (Figure 1), but there is a possibility that the double vaccinations these birds received (unsuccessful initial vaccination followed eight weeks later with a successful vaccination and “take”) may have provided additional protection against challenge with wild virus. Lesions among four of five vaccinated birds were small with healing times ranging from 12-14 days after challenge with Pox Variant 1. The wing lesion on the remaining vaccinated bird was moderate in size with a 6-7 week healing time, suggesting that the double vaccination had little or no cumulative effect. By contrast, lesions among unvaccinated Hawai`i `Amakihi were necrotic in all 5 birds, proliferative in four individuals, and led to eventual deaths of two Hawai`i `Amakihi, although effects on mortality were not statistically significant with our small sample sizes (Figure 18). We found that Fowlpox was only marginally infective to Hawai`i `Amakihi, leading to development of only minor lesions at the inoculation site in both vaccinated and unvaccinated birds. Similarly, the reciprocal experiments by Kim and Tripathy (2006b) provided evidence that both Nēnē (Branta sandvicensis) and Palila (Loxiodes

14

balleuili) isolates of the virus have only minimal infectivity to domestic chickens, but their precise relationship to Pox Variants 1 and 2 in our study are not clear since a different gene region was sequenced. These data provide additional support to the idea that domestic poultry and their associated Fowlpox infections may be of little threat to native forest birds and that major pox epidemics that were described near the end of the nineteenth century were associated with introduction of passerine Avipoxvirus from one or more non-native perching birds (Tripathy et al. 2000, Jarvi et al. 2008). Our data support limited experimental data from Jarvi et al. (2008) that Pox Variants 1 and 2 differ in pathogenicity in Hawai`i `Amakihi. Mortality among unvaccinated Hawai`i `Amakihi after challenge with Pox Variant 2 was 100%, with birds rapidly succumbing to infection within 30 days after challenge from extensive spread of the virus throughout epithelial tissue. Lesions in unvaccinated Hawai`i `Amakihi challenged with Pox Variant 1, by contrast, were more typical tumor-like swellings on the wing, beak, and feet, with only one bird exhibiting the disseminated, rapidly fatal, epithelial lesions that we observed in birds challenged with Pox Variant 2. Differences in survivorship between unvaccinated birds challenged with Pox Variant 1 and Pox Variant 2 were highly significant (P = 0.002, X2 = 9.496, df = 2) (Figures 10, 19). Effects on cumulative food consumption and weight change over time were most evident for birds challenged with Pox Variant 2, with an increase in weights within several days prior to death that corresponded to the development of extensive, edematous swelling of epithelial tissue in these birds. Similar epithelial lesions have not been reported in other experimental studies of pox virus in passerine hosts, including the recent report by Jarvi et al. (2008) that challenged Hawai`i `Amakihi by needle inoculation in the foot pad with Pox Variants 1 and 2. There are no reports of comparable lesions in wild birds with pox infections, and we have not seen similar lesions in native or non-native forest birds in almost 20 years of field work in Hawai`i. Given how unusual the lesions are plus their development in birds challenged with both pox variants, their pathogenesis may be related to both the high host susceptibility of Hawai`i `Amakihi to the virus and route of inoculation through the wing web. We found both extracellular viral inclusion bodies and mononuclear cells with viral inclusion bodies in subdermal tissue of these lesions which suggests that spread may have been through the lymphatic vessels in the subdermis or

15

movement of infected macrophages. While other studies have reported infection of macrophages with viral inclusion bodies of avian pox (Giddens et al. 1971, Shivaprasad et al. 2009), we did not find them in other tissues or organs and found no evidence of disseminated infections in epithelial cells of other organs (Shivaprasad et al. 2009). Based on our results, use of live attenuated vaccines for protecting native honeycreepers from avian pox is risky and these vaccines should be carefully assessed under controlled conditions to evaluate safety and efficacy prior to use in the field or in captive threatened or endangered forest birds. Given the potential that attenuated vaccine strains of the virus may revert in the wild and become more virulent, a preferable approach may be to develop effective subunit vaccines that can stimulate immunity to one or more wild variants of the virus without the risk that live viruses pose.

ACKNOWLEDGMENTS We thank the U.S. Geological Survey Wildlife and Terrestrial Resources and Science Support Programs for financial support, Bernard Rocha and Kawena Wise for providing daily care for Hawai`i `Amakihi and collecting data on food consumption, Nick Shema and the interns and technicians of the U.S. Geological Survey Palila Project for assistance with capture of wild Hawai`i `Amakihi on Mauna Kea and Kevin Brinck for invaluable advice about statistical analysis of the data. All animal challenges were conducted with approval of the University of Hawai`i Institutional Animal Care and Use Committee, Protocol 06-028.

16

LITERATURE CITED

Atkinson, C. T., K. L. Woods, R. J. Dusek, L. S. Sileo and W. M. Iko. 1995. Wildlife disease and conservation in Hawaii: pathogenicity of (Plasmodium relictum) in experimentally infected Iiwi (Vestiaria coccinea). Parasitology 111:S59-S69.

Atkinson, C.T. and D.A. LaPointe. 2009. Introduced avian diseases, climate change, and the future of Hawaiian honeycreepers. Journal of Avian Medicine and Surgery 23:53-63.

Docherty, D.E. and Slota, P.G. 1988. Use of Muscovy duck embryo fibroblasts for the isolation of viruses from wild birds. Journal of Tissue Culture Methods 11:165- 170.

Giddens, W. E. Jr., L. J. Swango, J. D. Henderson Jr., R. A. Lewis, D. S. Farner, A. Carlos, and W. C. Dolowy. 1971. Canary pox in sparrows and canaries (Fringillidae) and in Weavers (Ploceidae): pathology and host specificity of the virus. Veterinary Pathology 8:260-280.

Jarvi, S.I., D. Triglia, A. Giannoulis, M. Farias, K. Bianchi, and C.T. Atkinson. 2008. Diversity and virulence of Avipoxvirus in Hawaiian Forest Birds. Conservation Genetics 9:339-348.

Kim, T. and D.N. Tripathy. 2006a. Antigenic and genetic characterization of an avian poxvirus isolated from an endangered Hawaiian Goose (Branta sandvicensis). Avian Diseases 50:15-21.

Kim, T. and D.N. Tripathy. 2006b. Evaluation of pathogenicity of avian poxvirus isolates from endangered Hawaiian wild birds in chickens. Avian Diseases 50:288-291.

Kumar, S., K Tamura, I. Jakobsen, and M. Nei. 2000. MEGA: Molecular Evolutionary Genetics Analysis, Version 2.1. Arizona State University, Tempe, AZ.

Shivaprasad, H. L., T. Kim, D. Tripathy, P. R. Woolcock, and F. Uzal. 2009. Unusual pathology of canary poxvirus infection associated with high mortality in young and adult breeder canaries (Serinus canaria). Avian Pathology 38:311-316.

Smith, S.A., and G.J. Kotwal. 2002. Immune response to poxvirus infections in various animals. Critical Reviews in Microbiology 28:149-185.

Tripathy, D. N., W. M. Schnitzlein, P. J. Morris, D. L. Jannsen, J. K. Zuba, G. Massey, and C. T. Atkinson. 2000. Characterization of poxviruses from forest birds in Hawai`i. J. Wildl. Dis. 36: 225-230.

17

U.S. Fish and Wildlife Service. 2006. Revised recovery plan for Hawaiian forest birds. U.S. Fish and Wildlife Service, Region 1, Portland, OR.

VanderWerf, E. A. 2001. Distribution and potential impacts of avian poxlike lesions in ‘Elepaio at Hakalau Forest National Wildlife Refuge. Pg. 247-253. in J. M. Scott, S. Conant, and C. van Riper III (editors). Evolution, ecology, conservation and management of Hawaiian birds: a vanishing avifauna. Studies in Avian Biology 22. van Riper C. III. 1991. Parasite communities in wet and dry forest subpopulations of the Hawaii common amakihi. Pp. 140-153 in J. E. Loye and M. Zuk (editors). Bird- Parasite Interactions, Ecology, Evolution and Behavior. Oxford University Press, New York. van Riper C. III, S. G. van Riper, and W. R. Hansen. 2002. Epizootiology and effect of avian pox on Hawaiian forest birds. Auk 119:929-942. van Riper C. III, and D. J. Forrester. 2007. Avian pox. Pp. 131-176 in N.J. Thomas, D. B. Hunter, and C. T. Atkinson, (editors). Infectious Diseases of Wild Birds. Blackwell, Ames, IA.

Warner, R. E. 1968. The role of introduced diseases in the extinction of the endemic Hawaiian avifauna. Condor 70:101-120.

18

Figure 1. A neighbor-joining tree (Kimura 2-parameter corrected distances) showing relationships among pox virus sequences amplified from native and non-native birds and Poximmune C® (Vaccine C06). Three clusters are present – one associated with Fowlpox from domestic poultry (bottom, FP), one that includes sequence from Canarypox (Variant 1, V1) and one that is closely related, but distinct from Canarypox (Variant 2, V2). All sequences in the Variant 2 cluster were isolated from birds in Hawai`i. Tree was generated using vaccine sequence plus sequences reported in Figure 1 from Jarvi et al. (2008).

AY30309CANARY AY30308 SPARROW APAP14.1 MO 2003 (`Apapane) HAAM22.7 HI 1998 (Hawai`i `Amakihi) ALAL21.1 HI 2003 (Hawaiian Crow) AY30310 CURLEW VACCINE C06 V1 HAAM18.1 HI 2004 (Hawai`i `Amakihi) APAP16.1 HI 2003 (`Apapane) AY318871 ATCC CANARY 79 APAP16.2 HI 2003 (`Apapane) LAAL19.1 OA 2004 (Laysan Albatross) HAAM27.7 HI 2005 (Hawai`i `Amakihi) 100 ELEP1.1 HI 1900 (Hawai`i `Elepaio) HSFN24.1 HI 1998 (House Finch)

99 APAP5.1 HI 2002 (`Apapane) HAAMB.07 HI 2002 (Hawai`i `Amakihi) HAAM15.4 HI 2003 (Hawai`i `Amakihi) 65 V2 IIWI2.12 HI 2002 (`I`iwi) PALI17.2 HI 2002 (Palila) HSFN4.3 MA 2002 (House Finch) AY30311 LOVE BIRD 99 AY530303 PIGEON AY530305 OSTRICH AY530302 FOWLPOX 76 CHKN23.1 HI 1993 (Domestic Chicken, Hawai`i) AY530304 TURKEY CHKN1.1 HI 2001 (Domestic Chicken, Hawai`i) 98 CHKN23.2 HI 1993 (Domestic Chicken, Hawai`i) FP AJ005164 FOWLPOX M25781 FOWLPOX

0.05

19

Figure 2. Normal vaccine “take” in Hawai`i `Amakihi 321 that resolved by Day 34 PI.

Day 10

Day 21

Day 34

20

Figure 3. Duration of vaccine lesions in 15 Hawai`i `Amakihi that were vaccinated in the wing web with Poximmune C®. Vaccine “takes” in four birds (315, 305, 304, 302) developed into potentially life-threatening lesions that involved most of the wing. Lesions spread to other areas of the body in two birds, leading to loss of a leg and eventual death in one individual (`Amakihi 315). Five Hawai`i` `Amakihi (306, 316, 317, 318, 321, 78404) developed vaccine “takes” that healed within 3-6 weeks after initial vaccination. Six Hawai`i `Amakihi (308, 312, 313, 324, 327, 328) did not develop vaccine “takes” and were subsequently revaccinated eight weeks after the initial vaccination. Five of these six birds developed “takes” that healed within 4-7 weeks after the second vaccination. `Amakihi 313 (not shown), did not develop a swelling that would indicate successful vaccination and was dropped from the experiment.

328 327 324 312 308 315 305 304 302 78404

Band Number 321 318 317 316 306 0 20 40 60 80 100 120 140 Duration of Lesion (Days)

21

Figure 4. Necrotic lesion that developed on Hawai`i `Amakihi 304 after vaccination in the wing web with Poximmune C®. A large lesion developed at the inoculation site in the wing web (A, arrow) that subsequently healed and fell off (inset) with some scarring of the wing web (B, arrow).

A

Day 84 Day 93

B Day 97

22

Figure 5. Necrotic wing lesion (arrow) that developed on Hawai`i `Amakihi 305 after vaccination with Poximmune C®. The lesion eventually healed without leaving a scar.

Day 53

Day 105

23

Figure 6. Development of proliferative lesions on Hawai`i `Amakihi 302 after vaccination in the wing web with Poximmune C®. A large lesion developed at the inoculation site in the wing web (A, arrow) with subsequent spread to soft tissue surrounding the beak (B, arrow).

A Day 40

B Day 40

24

Figure 7. Development of proliferative, necrotic lesions (arrows) on Hawai`i `Amakihi 315 after vaccination with Poximmune C®. Lesions led to loss of function in one wing and eventual loss of the right leg.

Day 10 Day 34

Day 57 Day 78

Day 93 Day 120

25

Figure 8. Cumulative food consumption (top) and weight change (bottom) for Hawai`i `Amakihi vaccinated with Poximmune C®. Control Hawai`i `Amakihi (red) were sham- vaccinated with distilled water. Four birds experienced vaccine reversions and developed proliferative, life threatening lesions. Significant within-subject Treatment*Time interactions (P=0.007, F=1.735, df = 36), suggests that treatment had an effect on individual weight change over time. Note declines in weight among birds with vaccine reversions that became evident after Day 40 PI when lesions were most extensive.

2000

1500

1000

500 CUMULATIVE(ml) TOTAL

0 0 10 20 30 40 50 60

1

0

-1

WEIGHTCHANGE (g) Unvaccinated Control -2 Vaccine Reversion Vaccine Takes

0 10 20 30 40 50 60 DAYPI

26

Figure 9. Development of proliferative, necrotic lesions on Hawai`i `Amakihi 321 after challenge with Pox Variant 2. Lesions led to loss of circulation in the right leg (Day 50) and loss of the upper beak (Day 50).

Day 2 Day 9

Day 22 Day 31

Day 46 Day 46

Day 50 Day 50

27

Figure 10. Kaplan Meier Survival Plot of vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 2. A control group of unchallenged vaccinated Hawai`i `Amakihi was included for comparison. While mortality among vaccinated Hawai`i `Amakihi was significantly lower than unvaccinated birds (P = 0.002, X2 = 9.651, df = 1), mortality was still 60%.

1.0

0.8

0.6

0.4

SURVIVALFUNCTION 0.2

0.0 0 10 20 30 40 50 60 DAYPI Control, Not Challenged Unvaccinated, Challenged Vaccinated, Challenged

28

Figure 11. Cumulative food consumption (top) and weight change (bottom) among vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 2. Vaccinated control birds were not challenged with live virus. There were significant Time (P<0.0001, F=1385.366, df = 7) and Treatment*Time (P<0.0001, F=3.121, df = 14) interactions for weight change, indicating that there were significant day to day variations in weight change within subjects that were affected by treatment. Note increase in weight among unvaccinated Hawai`i `Amakihi between 20-30 Days PI, just prior to death.

2000

1500

1000

500 CUMULATIVE(ml) TOTAL

0 0 10 20 30 40 50 2

1

0

-1

WEIGHTCHANGE (g) Vaccinated Not Challenged Unvaccinated Challenged -2 Vaccinated Challenged

0 10 20 30 40 50 60 DAYPI

29

Figure 12. Hematoxylin and eosin-stained section of a wing lesion from Hawai`i `Amakihi 309 after vaccination and challenge with Pox Variant 2. Viral inclusion bodies (arrows) are evident as pink-staining inclusions within cells of the stratum germinativum (SG), immediately below the detached stratum corneum (SC) (inset). Necrotic areas and inflammatory infiltrates have extensively modified the underlying dermis and subdermis.

SC

SG

30

Figure 13. Wrinkled, edematous abdominal skin (upper inset) and hematoxylin and eosin-stained section of abdominal skin from Hawai`i `Amakihi 307 after challenge with Pox Variant 2. Viral inclusion bodies are evident as pink-staining inclusions (arrows) within cells of the stratum germinativum (SG). Necrotic areas and inflammatory infiltrates have extensively modified the dermis and subdermis.

SG

31

Figure 14. Hematoxylin and eosin stained section of the subdermal region of Hawai`i `Amakihi 331. A. Macrophage with intracytoplasmic viral inclusion (arrow). B. Extracellular viral particles (arrows).

A

B

32

Figure 15. Development of small lesion in Hawai`i `Amakihi 328 after vaccination with Poximmune C® and challenge with Pox Variant 1. Four of 5 vaccinated birds developed small, self limiting lesions (arrow) that healed completely between 5 and 19 days after challenge.

Day 0

Day 5

Day 19

33

Figure 16. Development of a moderate lesion in Hawai`i `Amakihi 312 after vaccination with Poximmune C® and challenge with Pox Variant 1. The wing lesion (arrows) in this bird did not completely heal until day 62 after challenge.

Day 0 Day 5

Day 40 Day 62

34

Figure 17. Unvaccinated control Hawai`i `Amakihi 323 inoculated with Pox Variant 1. Note extensive swelling of the lesion (arrows) and eventual spread to the beak prior to death at day 68 post infection.

Day 0 Day 5

Day 33 Day 40

Day 62 Day 62

35

Figure 18. Unvaccinated control Hawai`i `Amakihi 78481 inoculated with Pox Variant 1. Note swelling of the lesion (arrows) that became hemorrhagic by day 19 post inoculation. At necropsy on day 36 post-inoculation, the skin over most of the body was edematous, and appeared wrinkled and blistered.

Day 0 Day 12

Day 19 Day 36

36

Figure 19. Kaplan Meier Survival Plot of vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 1. Mortality among unvaccinated Hawai`i `Amakihi was 40%, but this difference was not significantly different from vaccinated birds for this small sample size (P = 0.134, X2 = 2.242, df = 1).

1.0

0.8

0.6

0.4

SURVIVALFUNCTION 0.2

0.0 0 10 20 30 40 50 60 70 80 90 DAYPI Unvaccinated, Challenged Vaccinated, Challenged

37

Figure 20. Cumulative food consumption (top) and weight change (bottom) for vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 1. Unchallenged birds were not available as a control group for this comparison. There were no significant between-group differences in either cumulative food consumption or weight change over time.

3000

2000

1000 CUMULATIVE(ml) TOTAL

0 10 20 30 40 50 60 70 80

2

1

0 WEIGHTCHANGE (g) -1 Unvaccinated Challenged Vaccinated Challenged

-2 0 10 20 30 40 50 60 70 80 90 DAYPI

38

Figure 21. Small lesion (arrow) in the wing web of Hawai`i `Amakihi 302 after challenge with Fowlpox virus. Both vaccinated and unvaccinated Hawai`i `Amakihi developed similar small lesions that quickly resolved within 12 days after challenge.

Day 0

Day 5

Day 12

39

Figure 22. Weight change and cumulative food consumption for vaccinated and unvaccinated Hawai`i `Amakihi challenged with Pox Variant 1. Unchallenged birds were not available as a control group for this comparison. There were no significant between- group differences in either cumulative food consumption or weight change over time.

3000

2000

1000 CUMULATIVE(ml) TOTAL

0 10 20 30 40 50 60 70 80 1

0

-1

-2

Unvaccinated Challenged WEIGHTCHANGE (g) -3 Vaccinated Challenged

0 10 20 30 40 50 60 70 80 90 DAYPI

40

Figure 23. Kaplan Meier Survival Plot of vaccinated and unvaccinated Hawai`i `Amakihi challenged with Fowlpox. One vaccinated bird died at Day 69 PI from causes that were not related to pox infection. There were no significant differences in survivorship between the two groups (P = 0.264, X2 = 1.250, df = 1).

1.0

0.8

0.6

0.4 SURVIVALFUNCTION 0.2

0.0 0 10 20 30 40 50 60 70 80 90 DAYPI Unvaccinated, Challenged Vaccinated, Challenged

41